Building a better ETF

September 28, 2015 01:00 PM

There’s ample opportunity for traders who want to speculate in the construction market. Exchange-traded funds (ETFs) and exchange-traded notes (ETNs) on construction materials include the United States Copper ETF (CPER), iPath Bloomberg Copper SubTR ETN (JJC), Guggenheim Timber ETF (CUT), iShares Global Timber & Forestry ETF (WOOD), iPath Bloomberg Aluminum SubTR ETN (JJU) and Market Vectors Steel Index ETF (SLX).

CUT and WOOD ETFs track different indexes, each based on approximately 30 securities connected with timber and lumber products with holdings in the United States and globally. For WOOD this is the S&P Global Timber & Forestry Index and for CUT the Beacon Global Timber Index.

The metals ETFs and ETNs are designed to provide investors with exposure to single asset classes: Copper, aluminum and/or steel. They reflect unleveraged market prices of futures contracts for the metals traded on COMEX and other metal futures exchanges. 

“Building materials” (below) shows the cumulative percentage price changes for the six securities for the period Jan. 2 through May 29, 2015. For most of this time period the majority of cumulative percentage price changes were below zero, so it is only from mid-April forward that the prices move from minus 10% to +5%. Only two of the six (WOOD and CUT) maintain a positive percentage change throughout, as the CUT ETF peaks at +10% and finishes ahead of the rest at +7%. 

Because the pairs included for copper and timber/lumber may present opportunities for spread trading, they are shown on separate charts in “Constructing returns,” (below). 

CPER, JJC and July 2015 copper futures’ cumulative percentage price changes are closely related throughout the period from January to May 2015. There are several days on which separations occur, but overall the market keeps the three price changes well under control. 

The second chart in “Construction returns“ shows a different picture from the copper price series. While the ETFs stay fairly close through the end of February 2015, the separation between CUT and WOOD widens increasingly through the first week in May. Meanwhile, the lumber May/September futures contract’s cumulative percentage price falls from early January, ending at -30%. 

The timber ETFs are based on a number of companies and products related to the timber, forestry and lumber industries — both foreign and domestic. Therefore, it is not surprising that their cumulative percentage price changes can be positive through most of the period while the lumber futures price is descending from zero to -30%. The separation between timber ETFs and related futures contracts is very different from the copper and aluminum ETFs, which show close correlation with futures price changes. 

Playing with spreads

The picture becomes more interesting when we look at the relationship between the CUT-WOOD spread in cumulative price changes and the comparative changes in lumber futures prices. “Spreading lumber” (below) shows how close these series are — from January to May — when the CUT-WOOD negative difference is multiplied by 5. 

The CUT-WOOD spread favors CUT, so it appears that either WOOD is forced down by the same elements that form the lumber futures contract, or that the CUT ETF is benefiting from the same elements that are causing the downward trend in lumber futures. From the first week in January to early May, lumber futures declined by a cumulative 30% while the advantage of CUT over WOOD increased by approximately 6%, or $12 million in market value on the funds’ worth of slightly more than $200 million each.  

“Metals magic” (below) shows dollar prices of the copper ETF (CPER) with aluminum ETF (JJU), and the copper ETN (JJC) with the steel ETF (SLX) from Jan. 2 to May 29, 2015. As potential spread trades, the two pairs have varied around each other as metals used in construction, and in both cases the exchange-traded securities are close enough in price to avoid using multipliers to adjust for size differences. During this five-month period there are numerous occasions for potential spread trades. The spread is necessary to control risk because the trade is not one that is based on direction of price change, but instead on the assumption that the two prices will converge after temporarily separating.

Copper options

The first chart in “Constructing returns” shows that the cumulative price changes for CPER, JJC and copper September 2015 futures are tightly coordinated. Having an idea of future price movements of any one of the three would enable forecasts for the other two as well. In this regard, it is helpful to find that the futures contract is so close in price movement.

On May 13, 2015, the LLP options pricing model based on 17 strike prices when September copper was priced at 2.9275 forecasted upper and lower breakeven prices of 3.255 and 2.760. The breakeven range of 11% and -9.4% from the current price may be applied to CPER and JJC. Based on the close-fitting price movements in “Metals magic” (above) the percentage changes for the upper and lower breakeven price range through September might be applied to steel SLX as well. 

The upper and lower breakeven prices are those that would result in a breakeven trade made on May 13, 2015, selling approximately two calls on September copper futures against one long futures contract.

“Construction materials” (below) includes the options price curves for the Home Construction ETF, September 2015 copper futures, the copper ETN (JJC) and the steel ETF (SLX) on May 29, 2015. With the Home Construction ETF in the middle, the curve for September 2015 copper futures is the lowest — showing the least volatility — while steel (SLX) and copper (JJC) have the two highest curves. The volatility measures of curve height as a percentage of the strike price are 5.92% for copper futures, 8.33% for Home Construction ETF, 11.06% for copper ETN (JJC) and 11.46% for steel ETF (SLX). 

The relatively high volatility measure for copper (JJC) compared with the low measure for copper futures conflicts with the chart, “Constructing returns.” That chart shows the three series in close coordination with little room for volatility differences. The implication of this conflict between market price comparisons and volatility shown by call options is that the copper ETF calls may be overpriced — giving a trader the opportunity to sell volatility with risk reduced by long futures. 

Any mispricing between copper ETFs and copper futures could be explained by the difference between high-volume trading of futures contracts vs. the low trading volume and small open interest of ETF and ETN options. For example, on June 3, 2015, ask prices on the copper ETF (JJC) December 2015 calls ranged from $0.10 to $2.15 for strike prices from $44 to $33 when the underlying was $32.58. Only one of the strikes merited an open interest of more than a dozen or so contracts, and this was the $33 strike with open interest of 608.

Comparison between the ask prices for December 2015 JJC calls and the prices predicted for the same options using the pricing formula computed for calls on December copper futures is shown on “JJC Options Comparison” (below). The ask prices form a smooth curve that rises with declining strike prices. At each strike there is a significant difference between the ask price and futures price. The implication is that JJC has greater price volatility — thus larger options value — than copper futures, whereas “Metals magic” showed that the volatilities should be approximately equal.

Today’s investment environment includes numerous ways to gain exposure to the same markets. Investigative traders can uncover low-risk divergences that can be exploited in creative ways.

About the Author

Paul Cretien is an investment analyst and financial case writer. His e-mail is